Eve Sweetser, 2003
I. Facing up to the problems with pigeon-holing — in general, as well as in Art.
Literary categories are strange beasts, to put things mildly. It would be quite depressing to seriously consider the number of trees killed, and the quantity of ink chemicals dispersed, in order to publish writings on topics such as “what was really the first novel,” or “what Romanticism (or Modernism, or…) really is.” Such authors normally seem to proceed by (1) presenting their new and different definition of the category in question (the Novel, Modernism) and then (2) showing that — surprise! — with this new definition, there’s a new answer to the question of whether some crucial work is Modernist, or a Novel. Explanations of what makes their definition more profitable than others seem optional; but even if they are present, they don’t change the basic circularity of this kind of argument. (How can you “disprove” someone else’s claim that X is a Modernist work, when you admit that you’re not using the same definition?) But the oddest part of these debates is that they continue as if sure that there is a yes or no answer to such categorization questions. They never seem to stop and wonder whether the problem lies in the fact that some entities may not fit the repertoire of standard categories well.
In general, humans are oddly rigid — and oddly flexible too — about categorization. In some respects, the rigidity makes sense: after all, physical objects like books and CDs can only be on one physical shelf at once. And it’s not a great deal of help in finding one, to be told that there’s a huge “unclassified” shelf area — category labels do help shorten search time. For this kind of purpose, I need to decide whether X is really a science fiction novel or a mystery, a romance novel or a “legit” historical fiction novel, if those are the available categories. On the other hand, digital databases can readily search directly by author or title without any intermediary genre, or access a given work via multiple genres if desired — as long as humans are flexible enough to tell them to. The problem lies not in our tools, but in ourselves.
I’m neither a fiction writer nor a literary scholar. I’m a linguist, specializing currently in the relationship between language and cognition. My daily bread includes current scholarship on human categorization — which goes a long way towards explaining how categories are formed, how they change, and why people are rigid about them (and sometimes flexible instead).
Before I get into summarizing some of the categorization research, let me add that my professional empathy is entirely with the Interstitial Arts movement. Academics can be worse about “boxes” than non-academic editors (IMHO, as they say!). I happen to work for the Linguistics Department, the Cognitive Science Program, and the Celtic Studies Program at UC-Berkeley. I’m lucky to be at Berkeley, where all these things can actually happen; but you try getting anyone to claim Celtic Studies as theirs. And most Linguistics Departments don’t teach courses on metaphor at all (I do so regularly). To make matters worse, I also work on language-accompanying gesture — which nobody thinks is inside their box. And my Celtic work has been largely on early Welsh poetic texts; at last count I was the only person on the large Berkeley campus who was bothered by the conflict between the CogSci Colloquium talks and the weekly Medieval Studies lunch. As I was saying to one of my students recently, we ought to start the Interstitial Sciences movement.
The views I’ll be presenting here are not just mine. But they are controversial in the scholarly community. They belong to a school of scholarship which treats cognition as essentially embodied. That is to say, our body’s neural system and dynamic perceptual interface with the world are essential in shaping not just emotion or aesthetics or cognition, but also rationality, logic, mathematics and the rest of the “objective” aspects of cognition. In such a worldview, imagination and reason, or art and everyday life, are not opposing categories, but deeply interwoven with each other. Metaphor is seen as a major building block of human thought. Yes, this is really different from most “objectivist” science. But it’s mainstream, even if it’s controversial; over the last twenty years it has become a major school of thought, distributed world-wide, though the Berkeley and San Diego Cognitive Science communities remain major centers for such work.
(Note: I will be restricting myself to the subject of categorization, and I will not be giving footnotes and references as I go. But interested readers can find a brief list of potentially interesting works from the Embodied Cognition school at the end of this essay; and those works contain thousands of references.)
II. Putting the “human” back into our understanding of human categories: it’s not just in art that objectivist category theory has been wrong.
When I say human categories, I mean human categories, as opposed to most of what gets talked about in Western philosophy, math, computer science, law, and so on. Scholars across various schools (not just Embodied Cognition folks) agree that the human neural system is a categorization system. It’s evolved to take in stimuli and group them according to similarities and differences that have proven useful to human animals and their ancestors. If we didn’t constantly categorize new stimuli relative to our extant category system, we’d be back to the condition of a newborn — most things would be brand-new every, time and we’d have to start over with identifying every new entity we encounter. It’s categorization — and I mean routinized, established, unconscious categorization — which lets us know that a chair is a chair, a floor is a floor, a book is a book, so that we can get on with life instead of needing to grab (and probably lick) every new object to see what it’s like.
The same is true of art and literature. If I didn’t have genre expectations — and general expectations based on previously encountered texts — I would not be a sophisticated reader, able to notice intertextuality, enjoy creativity, differentiate expected from unexpected elements, and helpfully fill in background from traditional expectations about a genre. Caroline Stevermer once told me that male readers of her novel Sorcery and Cecelia (co-written with Patricia Wrede) expressed enjoyment of the book’s wit and humor — but puzzlement over the fact that the authors made it so obvious, so soon, who was going to marry whom. To female readers, more familiar with the romance genre, the obviousness of Wrede and Stevermer’s heroes as matches for the heroines was part of the spoof on that genre. When you see the tall, dark, fascinating but arrogant guy, and sparks flying between him and the heroine, the ending should be predictable. If we didn’t have entrenched categories, we’d have nothing to play with, nothing to play off. It would all be starting over again, every time.
So entrenched categories are good. The problem comes when they don’t deal well with new input. The human system can — here’s the good news — be flexible enough to change when that happens.
Real human categories have fuzzy boundaries, and better and worse members. Most of us who were educated since New Math days know that set membership is a yes-or-no thing; any entity is a member of either set A or the complement set not-A. You can’t be more or less a member of A: all members are equal. But all the evidence from cognitive psychology, linguistics, and social sciences in general says that that’s not what most human categories are like.
Let’s take an example. Is it the case that the world is divided into red objects and non-red objects, and all we have to do is categorize an entity as one or the other? Well, nobody really thinks that maroon is as red as fire-engine red; but everyone thinks maroon is redder than, say, lime green — and even that it’s redder than royal purple. When do we stop saying something is red and start saying it’s purple? People who know both words can fight about that — recognizing full well that that color range is both reddish and purplish, but is not the best possible example of either category. This is a normal, not a weird, situation in human cognitive categorization.
Or how about the category pet? Is an ant as good an example of a pet as a cat or dog? People do keep ant farms. But most people would agree that if a motel said “pets allowed,” they might still not let you bring your ant farm — or your goat, your python, your chimpanzee — they might even object to your duck or your chicken. They were thinking of dogs, cats, gerbils and parrots — possibly iguanas at a stretch. Unlike Boolean sets, human categories quite normally have central and less central members. Art is a clear example of a fuzzy category. Ballet is art; is folk dancing? Are ballet classes “art”? A Faulkner novel is art; is a mystery story art? And so on. Literature is another such fuzzy category; for the moment I leave you to compute that one on your own.
Categories can overlap and/or include each other: boundaries don’t need to be exclusive. Something can be blue and red at once (in fact, that’s what purple is, a combination of visual response in the red and blue neural patterns). This is an area where humans often seem to be particularly persistent in denial, though. If something is a member of one category, people don’t like to call it simultaneously a member of another contrasting category. Shirt and jacket, or cup and bowl, are examples of cases where one object may sometimes be used as a member of either of two normally contrasting categories. Drawing an immediate parallel with art and literature, it should be obvious that there are plenty of cases where an entity belongs to multiple artistic genres simultaneously.
It’s OK to give two labels if the categories are not contrasting; for example, human is not in contrast with mammal at a single level of categorization, but mammal is rather a superordinate category of which human is a subclass. So nobody has trouble saying that she’s both a human and a mammal, or both a human and a woman. This brings up other issues, though; I wouldn’t call myself a mammal, a human, and a woman under exactly the same circumstances. Levels of categorization are important to construal: when I say “I’m a mammal,” you know I’m comparing myself not with (let’s say) male humans, but with fish or lizards.
It can also be OK if the categories are simply orthogonal to each other; for example, saying I’m a professor, an amateur painter, and a woman. All of these categorizations may say things about me, but what they say is largely independent, though all are subcategorizations of humanity. Of course, orthogonality of categorization is as culture-dependent as the categories themselves; if women couldn’t be professors, then professors would just be a subclass of men, rather than of humans.
Returning to art (and specifically to literature, within art), there are of course literary categories with serious potential for contrast and conflict; some of them have quite fuzzy boundaries. Comedy and tragedy, for example: we’re sure that the central examples are very different, but there’s clearly fuzzy ground in between (and plenty of ink has been spent on that too — is Measure for Measure a comedy or a tragedy?). But in the literary and artistic world, as we know, it can even be a problem for a work to fit into two relatively orthogonal categories. There’s no obvious problem in fulfilling the usual demands of category membership in both mystery and science fiction simultaneously. Asimov — who pretty much set out to prove that science fiction was compatible with nearly every other genre — made early sci-fi history with Lije Bailey, and others have followed. But should a mystery set in an imagined world be marketed as a mystery, or as a science fiction or fantasy novel? You might think the answer would be simply, “yes.” But of course, in fact the answer depends on questions like whether the author has previously written mysteries or science fiction, which genre her editor can market best, and so on.
That is to say, although the literary categories mystery and science fiction, or romance and fantasy, may be (for the author) as compatible and non-competing categories as woman and teacher, the corresponding marketing categories, reader community categories, library categories, and literary review categories are complementary. Did Lije Bailey and R. Daneel mysteries reach those mystery shelves? — not very often, I imagine. And for better or worse, I haven’t seen Sorcery and Cecelia on romance shelves, though I admit I don’t track those as carefully as the fantasy shelves. It’s kind of like a culture that has declared women can’t be teachers; there’s category complementarity imposed by causes other than the internal category structure itself. And it’s not simple; if women can’t be teachers, women won’t get educated, so there’ll be no women qualified to be teachers, so it’ll be correct to say that the teacher candidate pool is men. If two separate communities with different reading experiences and preferences have been set up for mystery and science fiction, those communities aren’t instantly unified by the book that happens to belong to both genres. And yes, in academia I’ve been asked whether I wanted a book to reach a linguistics audience or a cognitive science audience; it is still not easy for an interdisciplinary author to get the right response to the answer, “yes” in such a situation.
Of course, it’s my opinion that we shouldn’t give up in these situations; editors and publishers are often too accepting of such boundaries. Often there’s more community overlap than they’re willing to admit; mystery readers may be a community non-identical to the fantasy/sci-fi reading community, but they’re by no means complementary. (The same is true of linguistics and cognitive science.)
III. Prototypes and membership criteria: Why you can’t just list necessary and sufficient conditions for membership in the category Art.
Real human categories are formed around “prototypes.” Boundaries are important to human categorization, but much of the real cognitive action is at the center. Judgments about membership are often as much based on perceived similarity to some idea of what a really good member of the set is like, as on careful consideration of precise boundaries. Consider the pet example. Cats and dogs are so central to American experience of pets that it would be really useful to tell the motel owner something about how your exotic pet resembles them, if you want to get the pet into the motel with you. (“My iguana is no bigger than a cat, and likes to sleep all day on a hot rock.”)
Some prototypes are based on actual individual salient instances of the category; for example, if you were brought up with a pet iguana, iguanas might well be more central to your pet category than to some other people’s — though you’d probably still be aware that that’s not true for everyone. Frequently, prototypes are based on information about a range of clear central instances of the category. For example, your prototype of car might include the average size of the central members of the category you had encountered. That might differ significantly depending on whether you’re American or European, on what generation you belong to (and therefore when you formed this category), and all kinds of other things. The prototype of the category art, for many Americans, seems centered on painting and sculpture (not pottery, which would rate high in East Asia or in some Native American communities), music (but classical, not folk or pop), and literature (well, the prototypical literature — i.e., poetry and “artistically written” prose fiction). Note that we don’t seem to have just one single prototype of art; although if we had to pick the most salient one, it might be the painting one. But it’s generally true that there are some central examples we all strongly agree on; the center(s) of the category art is a lot easier to get consensus on than the boundaries.
Categories don’t need to have some single invariant set of membership criteria. When you take American college students and ask them (as I often do, in my Cognitive Linguistics class) what the definition of art is, they come up with completely non-intersecting sets of criteria. Some of them hold hard to the view that aesthetic value is crucial; some are more interested in whether the impulse of the creator was personal expression; others are viewer-response theorists, and say art happens only if a response is intended (or even if a response actually happens) in the experiencer. Other relevant criteria include non-functionality (Americans see prototypical art as being non-functional, “just” aesthetic — so pottery and architecture aren’t core examples of art), special skills needed for the creation or performance (folk dance is less “artistic” than ballet because it’s the kind of thing anyone can do), and of course particular traditional media (painting, music, dance, the written or spoken word, and so on) and culturally-valued settings (museums, concert halls) which are traditional places for experiencer interaction with art. Linked to the special skills requirement, uniqueness or rarity — and direct connection to the person with the special skills — is also important; mass-made objects, even if designed by a really good designer, don’t count as being “art.” A hand-made original by the same designer might count.
We might add things like elite status; museums seem to be full of painted and woven objects made by elite creators for elite classes, more than they are full of painted and woven objects made by folk for their own everyday use. For some, we might even add spiritual or cultural “value” — if not moral value — as a criterion; others would vigorously deny this. And non-profit status seems to help some, too; the better someone “sells,” the less pure we seem to feel the artistic motivation is. (Sure, we are happy to know that Yo-Yo Ma is not starving, and we think he’s a truly great artist; but even here, it does help that we know he doesn’t make the kind of money that the most successful pop musical stars do.) This is probably related to the non-functional criterion mentioned above: art should be done “just” because it’s creative, or beautiful, or thought-provoking, or culturally relevant — not because the artist is paid, or because the experiencer or consumer gets useful things like a building, a blanket, or teapots and cups.
A categorization theorist would say none of these definitions is “wrong.” And in fact, the students usually know that they all agree more on the best examples of the category art than on the boundaries of the category, as mentioned above. The Anglo-American prototype of art, one might say, has all (or most of) the features the students bring up; where they vary is in the importance they attach to some of those features, in the absence of others, in establishing category membership at the periphery. Art theorists have been pushing at this category forever in ways which look predictable to a category theorist: objets trouves, especially that famous Duchamp urinal, were put forward precisely to argue for the view that “beauty” and specially skilled creation were not the “essence” of art — rather, art was making people think and react (in this case, making them think about their definition of art). The 20th century was a period of huge rethinking of this category, both on its own and in relation to other categories like craft and design.
Categories are not just “feature-based”; they’re strongly interaction-based. Even at what seems like the most basic level of human categorization, categorizing physical objects, human interactional “affordances” are central to how we divide up the world. Someone who didn’t know a beanbag chair is meant to be sat on would not categorize it with chairs. Linguists have found that people’s use of labels like cup and bowl depend partly on shape and size, but also partly on what the dish in question is being used for (does it hold food or drink?). Calling something a stool or a side-table might depend similarly on whether it is being used for sitting or as a convenient surface for putting objects on. We noted above that this is true of art and literature as well. Things brought into museums and concert halls become art. Things taught in literature classes become literature. Experiencers, and cultural evaluation, are part of the frames of art and literature.
And here’s another place to bring up the “imposed” complementarity of genres like romance and fantasy. Each of those genres has not only internal structure, but external affordances for a reader/user community. If the communities are separate and socially differentiated (as they clearly are), then those external affordances and uses of the two genres are separate, no matter how compatible the internal structures of the genre categories may be. This is equally true of physical objects. A bone-china saucer may make a great ash-tray or cat-food dish — it has all the right purely physical affordances; but the owner may not want it to be used for those purposes. The same owner might be equally unwilling to use a cat-food dish as a tea saucer, even if it had been safely through the dishwasher and was shaped appropriately for use as a saucer.
IV. “Essentially contested” concepts and categories — such as Art — and how category structure affects reasoning about everyday life.
Art is a classic example of another now-recognized phenomenon in category theory — as is clear from my students’ discussion, it’s a contested category. It is, in fact, an “essentially contested” category; that is, we live with the knowledge that speakers don’t agree about this category in really basic ways.
We should be proud of this in one sense; essentially contested categories never happen in unimportant areas of cognition and culture. Contestation of the label art is a tribute to the cultural importance and status of the category. Other cultural institutions reflect this; in the days when the National Endowment for the Arts had money to give, presumably you had to be doing something that counted as art to be funded. A piece of functional pottery which counts as “artistic” costs a lot more money (even in America!) than one which is not “artistic.”
Categories are relative to context, not absolute. Just as the same person may be a Liberal in Texas and a Conservative in Berkeley, or (in my case) a Flaming Radical in Texas and a Liberal in Berkeley, so a great many categories are constituted differently depending on context. Tell me that someone is a wife or a husband, and I’ll want to know what culture that’s in, and what their marriage customs are, before I really know what you mean, even though I’ll know that wife and husband probably refer to monogamous heterosexual legally sanctioned sexual relationships. (Here we have translation problems. There are some terms, it seems, in nearly every culture, which refer to such legally sanctioned relationships; and they get translated into English as wife and husband, even if the culture they’re being translated from is polygamous, for example.)
Similarly, it’s hard to imagine what it really meant to be a Romantic without living in the right part of the 19th century, and rebelling against Rationalism. I recently saw a new comic series called Last Kiss Comics, which is produced by taking 1950′s romantic comics (one series was called First Kiss Comics) and putting different words in the speech-bubbles, while preserving the pictures. The 1950′s romance genre will obviously never mean quite the same thing to me that it does to someone who lived the 1950′s as an adult, as my mother did.
Prototype structure of categories is important in reasoning. Experimental work has found that people reason from knowledge about central (prototypical) members of a category to infer things about more peripheral members, but not the other way around. Here’s an example: for most urban Americans, a robin is pretty close to the prototype for the category bird. If you ask such people whether an ibis or an ostrich is likely to catch a disease from a robin, they rate that as more likely than that a robin will catch a disease from an ibis or an ostrich.
Tragically, this kind of reasoning probably played a role in keeping the American public from early response to the dangers of AIDS. If your prototype of human is white and heterosexual, then you subconsciously reason that diseases which afflict white heterosexuals could surely spread to gay people and to African or African-American people — but you don’t as readily worry about “gay” or “African” diseases spreading to affect the straight white population. (I won’t even go into whether inner-city intravenous-drug users get counted as full human beings, or the assumed connections between race and drug abuse.) In short, the scarily-wrong labeling of AIDS as a “gay” disease or an “African” disease came directly out of general human strategies of category-structure-based reasoning, in combination with some noxious construal of the category human.
Reasoning about art has similar problems. It’s a normal situation for there to be community-specific genres of plastic arts, performing arts, literature, you-name-it. Tax money (predictably) supports genres supported by the more powerful sub-communities. Strangely, this support is sometimes given in the apparent belief that classical music (for example) is accessible and relevant to everyone, while bluegrass or Cajun music or punk rock belongs to a sub-group. I’m not trashing classical music here; I not only love it, I feel it’s under-supported, like arts in general in this country. I’m just making some observations. The more we question the Canon, the more difficult it is to say what it means to support culturally-meaningful artistic endeavor.
V. Why this all matters: (cross)cultural understanding and meaning, in Art and in Life.
We can’t stop categorizing; stereotypes are therefore inevitable. My students are always eager to point out that a lot of the conventional category prototypes are “stereotypes.” The uncomfortable moment in the classroom comes when I point out that you can’t just ditch stereotypes, throw out all the “bad” categories and somehow keep the “good” ones. Just as harmful or unpleasant mutations (and cancer) are natural and inevitable aspects of the balance between flexibility and rigidity in cell reproduction, similarly it is inevitable that the constant ongoing process of human categorization will result in some harmful categorizations — or old categorizations that were helpful but are really unproductive in a new context. The best we can do is to make ourselves as aware as possible of our category structures, and notice when they are not working productively for us.
Categories are flexible as well as rigid. The fact that we are not making Boolean sets, but fuzzy-bounded, prototype-centered categories, means that mechanisms for change and variation in category structure are built right into our categorization processes. If you get different input (see different cars in your environment, e.g.) you will have a different prototype; and indeed, as the pool of cars you see changes over time, your prototype changes too. (Don’t cars from 30 years ago look funny to you now, although if you’re old enough to remember them, they didn’t look funny then?) If we run into a newly invented kind of object, the fuzziness of the boundaries of old categories may permit us to debate whether it should be included in one of them.
This of course does not mean that humans can’t be rigid and unproductive in their use of categories. I began by mentioning some salient cases of rigidity. The basic fact is that we need to categorize everything in order to deal with it. Feeling uncertain about classification, or disagreeing about it, are therefore often uncomfortable situations.
In many cases, there may not be only one “right” categorization in objective terms; but there are much more and less “right” ones in human terms. Sure, just as there’s no single right reading of a work of art, there’s not necessarily one right categorization either. But there are wrong ones, ones that don’t resonate with any aspect of human experience. We’ll never have access to objective criteria for categorization, but we do have common access to human neural structure, which is the same kind of neural structure that created the art we’re experiencing and categorizing.
I once discussed this issue with a literature teacher who said she had recently been dealing with an intelligent student who claimed that (since modernism was over and postmodernism at hand) any reading was as valid as any other. The teacher felt that this attitude was getting in his way as a reader; he was facilely making up random readings from his own thoughts and interests, without carefully attending to what readings were more plausibly connected to the text. One day in class, she pointed out the window at the empty sky and said, “Look!! There’s a big yellow blimp out there!” He replied, dismissively, “No, there’s not.” She repeated her claim. He got it; and thereafter, he worked harder at his readings and listened better to her criticisms. Literary readings are a lot less simple to judge than claims about whether or not a large yellow object is in direct view or not. But humans don’t vary randomly and infinitely in their understanding of texts — though they do vary widely, both within and across cultures. And we should expect some shared ground, since humans make texts, and humans read them (despite all theories about the Death of the Author) knowing that they were made by humans. Martians, or Elves (for all I know) might have very different intentional and interpretive capacities.
I don’t mean by this to minimize the importance of cultural variability, or even personal taste, as factors in artistic creation and response. I do mean that when two people A and B are taking part in the same cultural frame, it would simply be silly to ignore the role of that common frame in interpreting what A’s art means to B. A and B take it for granted. And on a cross-cultural scale, it would be silly to ignore what we know about universals of human perception, interaction, etc., in examining the limits on variability in artistic response — just as silly as ignoring the role of cultural difference in bringing about variation in response.
More ideas on how this might affect the artistic world? Well, I personally see artistic genres as fuzzy-edged, prototype-structured, and often contested categories: in short, classic examples of human categorization. Knowing that’s the case helps keep me from getting into those circular discussions about what really counts as a novel; that’s at least some benefit. It also helps me to understand that even the folks involved in those discussions are — in their way — exploring problematic boundaries, even if I think they’re wrong that a good solution has to necessarily find a single boundary.
Awareness of this situation also helps one comment on it, manipulate it, and explore it at the meta-level. Consider a piano composition consisting of opening the piano and sitting at it for a measured number of minutes without playing it (an actual, and highly controversial, John Cage piece). This piece may or may not be music: that depends on whose category of music we consult. Further — like the Duchamp urinal — it may or may not be art, depending again on whose category of art we consult; it might rate high in the opinion of those for whom art’s most central characteristic is making people think and react in new ways, but low in the view of those who think that music in particular must involve sound, or that art in general should involve special skills like piano-playing. At any rate, it certainly does explore the category music creatively at the meta-level. It toys with audience expectations, and what happens when they are not met. It makes us wonder what musical performance is for, and how much of it is constituted by the sound, as opposed to the context. It may bring musical memories or images to audience members’ minds, thus creating musical experiences (perhaps different ones for different audience members) without making physical sounds. It creates social awareness of concert frames for musical performance and how they are constituted. It’s clearly not just academic meta-analysis of music; it’s more like music-based performance art, perhaps? And so on. It’s more interesting to notice all this than to fight over whether it’s “really” music, according to different warring definitions of music.
A raised consciousness about category structure, and the basis of human cognition in embodiment, may also be a help in some other standard artistic dilemmas. I can’t count the times I’ve heard painters and musicians say that maybe literature can’t transcend cultural boundaries (darn those different languages, and that translation problem!), but painting and sculpture and dance and music are pan-human. The last time I heard this claim, it was phrased as, “when you paint, you paint in Human, not in English or in German.” Of course, that claim is far too simplistic. It’s not the case that, let’s say, a Western European or American can look at a Native Australian painting and automatically “get” what it means to the culture which produced it (though they might indeed find beauty in it), any more than a Native Australian can look at the work in a contemporary New York or London gallery and fully “get” it without having learned something about the relevant artistic culture. In a sense, we can be deluded by the lack of an obvious language barrier into thinking that nonlinguistic art is much more universal than it is — especially if we’re part of a dominant culture, and the rest of the world is at least partially tuned to our wavelength (it was a British artist who actually said we paint in Human). But I know how much education it takes to understand a Van Eyck seriously, as a modern viewer; should we assume that all present-day art is so much “easier?”
On the other hand, part of what the artist who said that meant, in context, was that his own art specifically was intended to play with aspects of visual perception — some of which might well be construed as generalizable across cultures, though perhaps not quite as simply as he hoped. (Vision science does seem to indicate that, predictably, the structure of visual perception and much of basic visual categorization is shared cross-culturally.) That kind of generalizability is a relief to think about, if you want to imagine that we can actually (by working at it) experience art coming from other cultures, other places, and other times — or even, perhaps, just from other people in our own culture who have very different points of view from our own. There may be aspects of our already-entrenched category structure that we can successfully exploit, even while also having to learn a lot of new categories to understand a new cultural context. Let’s hope this is so, because cultural variation is a slippery slope in art. When August Wilson says that no actor should play a role of a different race than his or her own, we know there are excellent reasons to avoid more white Othellos in blackface; but there’s also the troubling feeling that we could extend this ban further to keep ourselves from ever playing (or reading?) Shakespeare or Moliere at all, since we aren’t members of the authors’ cultures. So, awareness of both the generally human basis and the particular cultural bases for our categories may be a big help in understanding what our art is really doing, and where gaps in cultural understanding are likely to arise.
And ideally, of course, understanding categorization helps one let go of category expectations, or open oneself to a wider range of possibilities, both as a consumer and as a creator of art.
VI. Suggestions of interesting reading on these subjects.
The few works I present here are by some of the major figures writing on Embodied Cognition. They are chosen partly as being salient and central examples; and partly as being accessible. I’ve left out things which are intended for narrower, more specifically trained audiences, in favor of ones directed at broader interdisciplinary audiences (references to the former works, including some of my own publications, are readily to be found in the latter, however). I’ve also left out articles in favor of books; many of these books are readily available in paperback on Amazon. By and large, these folks write well — even the most academic prose on this list is straightforward and unstuffy, while the best is downright…artistic.
On categories formed by “embodied minds:” George Lakoff is one of the founders of Cognitive Linguistics and of what might be called Embodied Cognitive Science. His Women, Fire and Dangerous Things: What Categories Reveal about the Mind (University of Chicago Press, 1987) is not new, but not yet superseded; and it is far broader — also far more interesting reading — than some of the attempts at superseding it. It covers the earlier literature on categorization in detail, so you’ll find a flock of added citations there. And it makes some of the arguments I have not made here, about why we can’t reduce human categories to logical ones. Among the literatures reviewed are the psychological literature on category processing, including Eleanor Rosch’s influential work on prototypes and basic level categories; the psychological and anthropological literature on color terms, including Brent Berlin and Paul Kay’s World Color Survey; work on cross-cultural variability (e.g. Whorf); and a range of philosophical and linguistic work at large.
Mark Johnson’s The Body in the Mind: The Bodily Basis of Meaning, Imagination, and Reason (University of Chicago Press, 1987). Johnson is a philosopher; this book tackles the ways in which a human’s bodily experience of the world constitutes a basis for meaning, imagination, and rationality. There are some fascinating discussions of artistic examples (in particular, one of balance in sculpture).
On embodied cognition, imagination, culture and creativity in general: Lakoff and Johnson’s joint Metaphors We Live By (University of Chicago Press, 1980) and Philosophy in the Flesh: The Embodied Mind and Its Challenge to Western Thought (Basic Books, 1999). The first is a highly readable work, which was just a beginning to the project it embodies; in a way, it’s an extended essay. It has been hugely influential, and translated into a myriad of languages. The second is the major pull-their-work-together book they couldn’t have written back in 1980; but, fat tome though it is, it is not written for a narrow scholarly audience in a technical discipline.
Mark Turner’s work in general. Turner’s a crossover figure, with a background in math and computer science before he took a Ph.D. in English; he’s now a Professor of both English and Cognitive Science at the University of Maryland. (He also happens to be married to fantasy author Megan Whalen Turner. Incidentally, both Johnson and Lakoff are married to artists; there is definitely something going on here.) One of the uniting themes of his work is the need for both literary analysts and cognitive scientists to see texts as among the most complex and interesting products of human cognition. You might try Reading Minds: The Study of English in the Age of Cognitive Science (Princeton University Press, 1991) or The Literary Mind: The Origins of Thought and Language (Oxford University Press, 1996), or Cognitive Dimensions of Social Science: The Way We Think About Politics, Economics, Law, and Society (Oxford University Press, 2001). His early book Death is the Mother of Beauty: Mind, Metaphor, Criticism (University of Chicago Press, 1987) is specifically on metaphor; the later ones are cognitive approaches to language, literature, and social science. Lakoff and Turner coauthored a delightful volume called More Than Cool Reason: A Field Guide to Poetic Metaphor (University of Chicago Press, 1989), which is exactly what it claims to be. Very generally accessible as well as insightful.
Turner’s recent book coauthored with Gilles Fauconnier is entitled The Way We Think: Conceptual Blending and the Mind’s Hidden Complexities (Basic Books, 2002). It’s a highly readable work, intended for people who are not experts in some particular field, because it’s intended to reach a highly interdisciplinary audience. It’s about strategies of conceptual integration (or blending) as the basis for human cognition, and as an explanation for how humans have art, language, and culture in a way that our closest non-human relatives don’t. It’s controversial, accessible, and state-of-the-art.
For added fun in the interface between culture and cognition, check out the Anthropology shelves and try Edwin Hutchins’ Cognition in the Wild (MIT Press, 1995) or Bradd Shore’s Culture in Mind: Cognition, Culture, and the Problem of Meaning (Oxford University Press, 1996). As ever, real data turn out to be even wilder and cooler than nearly all fiction. These are the Margaret Meads of their generation, and their interest is cultural cognition in context.
And for anyone interested in specific discussion of how mathematical cognition is based in bodily experience and in metaphor, try George Lakoff and Rafael Nunez’ Where Mathematics Comes From: How the Embodied Mind Brings Mathematics into Being (Basic Books, 2000). Both Lakoff and Nunez are polymaths, and Lakoff here returns to an early academic interest in math and logic.
And for those with interest in the neuroscientific basis for this approach to cognition, try out Antonio Damasio’s readable and fascinating The Feeling of What Happens: Body and Emotion in the Making of Consciousness (Harcourt, Brace, 1999). Antonio Damasio and Hanna Damasio are leaders of the current neuroscience community, and he eloquently here explains his certainty that neuroscience does not let us separate rationality from emotion. Or try the fascinating reinterpretation of modern neuroscience in a Buddhist philosophical framework (Varela was a highly regarded neuroscientist, and Rosch is both a top experimental cognitive psychologist and a practicing Buddhist), in Francisco Varela, Evan Thompson and Eleanor Rosch’s The Embodied Mind: Cognitive Science and Human Experience (MIT Press, 1991).
About the Author
Eve Sweetser was brought up in Minneapolis by teacher parents and now teaches Linguistics, Cognitive Science and Celtic Studies at the University of California at Berkeley, where she’s Director of the undergrad Cognitive Science Program.
“What I love most in teaching,” she says, “is seeing students step outside disciplinary boxes and make new connections that their seniors (including yours truly) have been too rigid to make. And what I love most about Berkeley is that it’s one of the world’s true Borderlands, intellectually, culturally, ethnically, and artistically.
“As a scholar, my world revolves around the ways in which language and thought and culture are interwoven. This isn’t just about literary language: in fact, I live in a world where metaphor is finally recognized (it’s always been there) as a major building block of scientific thought and everyday cognition, not of literary language alone. But great literature is one of the most marvelous products of human cognition — and is also (forgive me) a rich, neglected data source for cognitive science. So whether or not all of my less interstitial colleagues understand, my work on metaphors in early Welsh poetry is part of my cognitive science work. I make students in my metaphor class read The Phantom Tollbooth, or Marge Piercy, or Micheal O’Siadhail — terra incognita to the more computational CogSci majors. (OK, why not analyze stuff I love? Plus, it’s a real kick to watch a smart macho computer nerd suddenly catch on to an Emily Dickinson poem for the first time.)
“I’d like to bring the literary and the cognitive-scientific worlds closer — to my mind, they need each other deeply. As with most dialogues across boundaries, the view from the other side can be a huge revelation about one’s own side. And that’s before we consider the possibilities of actually breaking down fences.”
Eve Sweetser can be reached at email@example.com.